Embedded Analytics & BI
You're paying six figures for BI nobody opens.
Tableau, Power BI and Looker bills keep climbing. Adoption keeps falling. The duty manager, the charge nurse, the store lead, the people who actually move the numbers, never log in. We rebuild your analytics on Apache Superset and ClickHouse, embedded inside the apps your team already uses.
SearchFIT.ai The problem
Sound familiar?
Every mid-market operator we meet, in Health, Hospitality, Retail and Logistics, describes the same five symptoms.
Our Tableau, Power BI or Looker renewal just landed and it's gone up 40 to 100 percent. Most seats are unused.
Our PMS, EMR or WMS has no real analytics inside it. Staff still run on spreadsheets emailed at 11pm.
Our analysts spend three days a week stitching reports nobody opens, and one day fielding 'just one more chart' requests.
Dashboards take 30 seconds to load. The exec team gave up asking for new ones two quarters ago.
We have data in five places: Postgres, the PMS, Salesforce, Stripe, a CSV nobody owns. There is no single source of truth.
What broken BI is actually costing you.
It isn't the licence fee. It's everything that doesn't happen because the data never reaches the floor.
Per year on legacy BI seats with single-digit weekly active usage.
Of analyst time burned on report plumbing instead of decisions.
Lag between an event happening and a manager seeing it on a dashboard.
Margin lost to over-staffing, no-shows, stockouts and dwell-time the floor never saw coming.
The fix
Put analytics where the work happens.
Standalone BI portals are where insights go to die. Embedded analytics flips it. Every chart lives inside the app your team already uses: the PMS for hotels, the EMR for clinicians, the WMS for logistics, the admin console for SaaS.
The duty manager sees occupancy by hour. The charge nurse sees length-of-stay against target. The store manager sees margin leakage before stocktake. No portal. No login. No "I'll ask the analyst on Monday."
We build that on Apache Superset for visualisation and ClickHouse for sub-second queries on a billion rows. Open-source, self-hostable, no per-seat tax. You own the stack and the data.
The escape route from legacy BI.
Most of our work starts as a rescue. The renewal is coming up. The contract has tripled. The CFO has had enough. We replatform without breaking the dashboards your business actually depends on.
From
Per-seat legacy BI
Six-figure renewals, slow extracts, single-digit adoption, locked-in semantic layer, and a long tail of "one analyst owns this dashboard" risk.
To
Embedded Superset + ClickHouse
Open-source, embedded into your product or ops apps, sub-second on operational data, semantic layer in code, and unlimited internal viewers.
Tuned for your vertical, not built from a blank Looker project.
We come in with the KPI definitions, semantic models and dashboard shapes that matter for your industry. You get to value in weeks, not quarters.
Health & Hospitals
For mid-market private hospital groups and clinic networks
The pain: bed occupancy reports land 24 hours late, theatre utilisation is a guess, and the RCM team finds revenue leakage at month-end instead of in-flight.
What we build: live occupancy, length-of-stay vs. target, theatre utilisation, readmission rates, RCM leakage and clinician productivity, embedded inside the EMR or admin console.
Hotel & Hospitality
For multi-site hotel operators and F&B groups
The pain: RevPAR and pace live in the PMS, labour cost lives in the rota tool, F&B lives in the POS, and nobody sees them on one screen until the GM meeting on Tuesday.
What we build: RevPAR, ADR, GOPPAR, channel mix, pace and pickup, F&B covers and average spend, labour cost percent, and housekeeping productivity, surfaced inside the PMS instead of a separate portal.
Retail & E-commerce
For mid-market retailers running stores and online
The pain: store managers learn about stockouts from customers, online margin is masked by promo, and merchandising decisions wait on a Monday extract.
What we build: basket economics, margin leakage, store-vs-online attribution, stockouts, returns velocity and promo lift, embedded into store ops apps and merchandising tools.
Logistics & Field Ops
For 3PLs, last-mile and field-service operators
The pain: dispatch quality is judged after the SLA's already breached, dwell time is a tribal-knowledge problem, and driver utilisation is reconciled monthly.
What we build: OTIF, dwell time, route productivity, dispatch quality, exception rates and driver utilisation, pushed into the WMS, dispatch console and driver apps in real time.
What we build.
Modern data warehouse
ClickHouse for operational analytics, Postgres for transactional truth, BigQuery or Snowflake when the workload demands it. Right tool, not religion.
Embedded dashboards
Apache Superset embedded into your product or internal apps with row-level security, white-label theming and SSO, so dashboards live where work happens.
Real-time event analytics
Streaming ingestion from Kafka, Kinesis or webhooks into ClickHouse for sub-second queries on operational data: bookings, orders, sensor events, vitals.
AI-ready semantic layer
A versioned, code-defined semantic model so dashboards, ad-hoc SQL and AI agents all answer the same questions the same way.
Self-serve data products
Curated datasets, certified metrics and a Superset workspace your analysts and ops leads can extend without breaking production.
Legacy BI migration
Inventory of existing Tableau, Power BI or Looker assets, criticality scoring, parallel-run replatform, and a clean cut-over before your renewal date.
Already shipping
In production across every PADISO app.
We didn't write a deck and call it a service. Every product we ship runs on this stack. Below are two of them.
Embedded Superset dashboards on top of a ClickHouse warehouse, surfacing fundraising and pipeline analytics inside the Capitaly.ai product.
SearchFIT.ai
Real-time event analytics on ClickHouse, with embedded Superset views powering search-quality and ranking insights inside the SearchFIT product.
A 12-week embedded analytics rollout.
From audit to cut-over in one quarter, with legacy BI running in parallel until the day you switch off the last seat.
Weeks 1 to 2
Audit & blueprint
Inventory existing BI assets, score by criticality and adoption, map sources, and design the target Superset and ClickHouse stack against your verticals' KPIs.
Weeks 3 to 6
Stack & ingestion
Stand up ClickHouse, define the semantic layer in code, wire in CDC from Postgres, MySQL or event streams, and back-load history.
Weeks 7 to 10
Embed & ship
Rebuild the dashboards that matter, embed Superset inside your product or ops apps with SSO and row-level security, and run in parallel with legacy BI.
Weeks 11 to 12
Cut-over & handover
Decommission legacy seats, hand over runbooks and training, and leave your team with a stack they can extend without us.
Who this is for.
Mid-market multi-site operators
Hospital groups, hotel chains, retailers and 3PLs running 20 to 500 sites. You've outgrown Power BI, can't justify Snowflake plus Looker, and need analytics on the floor instead of in a portal.
Post-Series-B SaaS scale-ups
You need to ship customer-facing analytics inside your product without rebuilding a charting library. Embedded Superset with white-labelled theming and per-tenant row-level security.
Enterprise teams trapped in legacy BI
Your Tableau or Cognos renewal just landed and it's untenable. We replatform critical dashboards onto Superset and ClickHouse before the contract clock runs out.
Stop paying for dashboards nobody opens.
Book a 30-minute BI audit. We'll map your current stack, your renewal exposure, and what an embedded Superset and ClickHouse rollout would look like, before your next renewal lands.
Book a 30-min BI audit